How we helped a cardiology practice add $2.6M in annual collections.

More patients were waiting, but routine work still needed the owner.

The owner wanted to add $2.2 million in annual collections and get his evenings back. New patients waited a median of 6.5 weeks, while he worked 66 hours a week. The practice had been trying to recruit another cardiologist, and its initial request to us was for AI that could replace a medical assistant.

But the nurse practitioners and device-clinic staff already working there still depended on him for routine decisions. Every device-review item went through his personal review, so adding support had left the same physician queue in place.

We began the engagement in early 2024 by studying where his time went. Before choosing what to automate, we needed to know which work actually required him and which work kept reaching him because no one else had clear instructions.

He could explain the decisions. His team had no written standard to use.

We paired a two-week time study with electronic health record activity and a review of device-clinic work. The study showed 4.6 hours of non-visit work per clinic day, excluding revenue-producing imaging and monitor reads.

To understand the device queue, we asked the physician to sort recent items using his own criteria. Of 910 items, 817 were routine. He could explain which ones needed him, but those distinctions had never been written down for the team to use.

That changed the solution. Replacing an assistant would leave the same decisions coming back to him. We first needed to capture his criteria, get his approval, and put them into the daily routing so routine work could reach the appropriate person without waiting for him.

We turned his criteria into work the team could complete.

We worked from his signed reports and explanations to write the routing rules. He approved them before they became instructions for the system. We then built a routing layer on the practice’s server that matched incoming device items to patient records and applied those rules.

Routine items went to the nurse practitioners, administrative work went to the coordinator, and flagged or uncertain items went to the physician. Each clinical item carried the relevant report and context, so the receiving clinician could review and sign it. Software routed the work, while clinicians interpreted and signed it.

The practice also needed to turn the released capacity into appointments and completed billing. We helped write the nurse practitioners’ procedures, and the practice changed how it scheduled and billed their work. Our billing check flagged signed device reports that hadn’t reached a claim, giving the billing manager a concrete gap to resolve.

Documentation was another demand on the physician’s evenings, so we designed a comparison of scribe options and the consent script. The practice chose and bought its scribe. That was part of the wider operating change, alongside the routing we built.

The routing had to match his judgment before the team relied on it.

We built a reference set from the physician’s past decisions so we could test whether the rules sent routine items and exceptions to the right people. We started with one device-data source, checked the results, and expanded only after the pilot met the agreed acceptance criteria.

The physician also audited a sample of cleared items each week. That gave the team a continuing check on its decisions after rollout, while the system reran the reference examples whenever rules changed.

A changed export exposed a check we had missed.

During rollout, a vendor changed an export column and 14 flagged transmissions went to the wrong queue. Our log review caught the error. We told the practice that morning, and the physician reviewed all 14 within a day. None required action.

We had checked the routing decisions without making an unexpected input format stop the process. We added that check, so a changed export halted routing and alerted the coordinator, then reran the reference tests at no charge. The correction addressed the input change that had caused the mistake.

We trained the team to change the rules without depending on us.

The practice would keep receiving new alert types after handover, so a fixed set of rules wasn’t enough. We gave the clinical lead, coordinator and IT partner operating instructions, trained backups, and taught them to run the tests before releasing a change.

After handover, the lead nurse practitioner and coordinator added a new alert rule. The physician approved it, and they reran the tests without us. Their team could now extend the routing while keeping clinical approval and checking in place.

More work cleared without him, and patients were seen sooner.

At the operating follow-up, 88% of device-review items cleared without the physician’s personal review, up from none. Median new-patient waiting time fell from 6.5 to 3.1 weeks, and his workweek fell from 66 to ~58 hours. Routine decisions no longer occupied the same share of his day, leaving more room for consultations and procedures.

Freed clinic time went into additional consultations and procedures, while the billing check helped the team find completed reports still missing claims. In a later follow-up, the owner confirmed that annual collections had increased by $2.6 million and attributed the full increase to our work. The increase surpassed the original $2.2 million growth goal.

See how we handled routing, review and handover.

The technical companion explains the data flow, clinical decisions, testing, export correction and measurement definitions.

The practice grew while the owner’s workload fell.

The practice had wanted another way to serve its patients without making the owner carry more work. We found the decisions that kept returning to him, turned his criteria into approved routing, and trained the team to maintain it. That gave the practice more capacity, shorter waits and a way to keep routine work moving.

The owner attributed $2.6 million in additional annual collections to our work. His shorter week mattered alongside that financial result: growth no longer required the same level of personal review of routine device work.

For a practice owner planning the next stage of growth, the useful question is which part of the work still depends on their personal attention, and whether changing it would help the practice reach its larger business goal.